On Tue, Jan 16, 2018 at 2:13 PM, Alan Camillo <a...@blueshift.com.br> wrote: > I was thinking about one of my client's problem. > He has a ranking that will be build on a Fluo application. But there are > some rules on the Fluo application to decide the ranking score of the data > based on the already ingested data. > > The problem comes when they decide to change the rule of the score. > What do we do to minimize the impact of the reprocessing all the data? > Are there someone facing something like that? Am I thinking in the right > way?
It really depends on the specifics. The most general answer would be that all of the data would need to be reprocessed. Depending on the particulars its possible only a subset of the data needs to be reprocessed. Personally I would start with the simplest approach and optimize as needed. > > Thanks again! > Alan Camillo > *BlueShift *I IT Director
